As the nervous system develops, neurons become specified to make precise patterns of synaptic connections. Some aspects of neuronal phenotype are likely to be determined by a cell's lineage and its position within the developing embryo. Chemical signals present in a neuron's target tissue provide another mechanisms for influencing a cell's developmental fate. Neural activity can also shape the final pattern of synaptic connections. A knowledge of how these three mechanisms operate is basic to an understanding of how nerve cells originally make and then refine and maintain operate is basic to an understanding of how nerve cells originally make and then refine and maintain their proper connections. Synaptic connections in the spinal cord made by primary sensory afferent fibers innervating muscle and skin provide an excellent system for studying this problem at the level of single, functionally identified cells. Sensory neurons innervating muscle spindles form monosynaptic connections with spinal motoneurons, whereas cutaneous sensory neurons do not. The peripheral targets of sensory neurons are important determinants of the relative numbers of muscle vs. cutaneous cell types. A major aim of this proposal is to determine if this effect is caused by the selective survival of pre-determined sensory neurons, or if peripheral targets have an instructive effect on sensory neuron phenotype. Sensory fibers in the frog can regenerate after they are interrupted in the dorsal root. Although they re-innervate motoneurons with a high degree of specificity, they do not re- establish their long-fiber tracts within the dorsal columns. A second aim of this proposal is to see if neurite-promoting factors that work in vitro will also stimulate axon regeneration within the white matter of the spinal cord. Such studies may help in achieving regeneration in human spinal cord after accidental injury. Finally, we propose to continue our studies of the mechanism of presynaptic inhibition of sensory-motor synapses in the amphibian spinal cord. Synaptic transmission between muscle spindle afferents and motoneurons, the pathways that mediates the simple stretch reflex, is inhibited by stimulation of other sensory afferents. Previous work has shown that a major portion of this inhibition is presynaptic and is mediated by the neural transmitter GABA. However, activation of both GABAA and GABAB receptors inhibits synaptic transmission in this pathway. Experiments using pharmacological agents to block each of these receptor types should enable us to determine which type is responsible for the inhibition observed physiologically.